ALT-ERGO(1)

📦
version 2.4.3 · 3 versions
, in fedora 44

🦋 ALT-ERGO(1) 🦋

Okay bestie, welcome to the Alt-ergo Manual 🌸 This is the ALT-ERGO(1) page, and it's the one that explains how to run the solver on a file. Let's get into it 💅

✨ NAME ✨

okay so the name is alt-ergo, and its whole job is to execute Alt-Ergo on the given file 🔥

✨ SYNOPSIS ✨

alt-ergo [OPTION]… [FILE]

✨ ARGUMENTS ✨

  • FILE — this is your source file 📄 It has to end in .ae, .smt2, or .psmt2. .mlw and .why are deprecated, so maybe don't start new stuff with those 🙈

✨ OPTIONS ✨

  • -d, --debug — flips on the debugging flag 🐛
  • -i FMT, --input=FMT — sets the default input format to FMT. It has to be one of native, altergo, alt-ergo, smtlib2, smt-lib2, or why3 ✨ This is super handy when the file extension can't tell Alt-Ergo which parser to use, like in JS mode, GUI mode, etc.
  • -m VAL, --model=VAL (absent=none) — experimental model support on labeled terms 🔬 VAL must be none, default, complete, or all. complete shows one full model, and all shows every model 💜
  • -o FMT, --output=FMT — controls the solver's output format 🎀 FMT must be native or smtlib. The native format prints Valid, Invalid, or I don't know, while the smtlib format prints unsat, sat, or unknown. If you don't pick one, Alt-Ergo uses its heuristics to figure out the right format based on the input format. And heads up: if you skip this flag, Alt-Ergo will set the output format based on the input file's extension 👀
  • -p, --pretty-output — prints output with formatting rules, headers, and colors 🌈
  • -r, --replay-used-context — replays using the axioms and predicates saved in a .used file 🔁
  • -S STEPS, --steps-bound=STEPS (absent=-1) — sets the maximum number of steps 🪜
  • -s, --save-used-context — saves the used axioms and predicates in a .used file 💾 This one implies --unsat-core, so it comes with that too 🫶
  • -t VAL, --timelimit=VAL — sets the time limit to VAL seconds ⏰ Heads up, this isn't supported on Windows 💀
  • -u, --unsat-core — experimental support for computing and printing unsat-cores 🧪
  • -v, --verbose — turns on verbose mode 📣

✨ EXECUTION OPTIONS ✨

  • --add-parser=VAL — registers a brand new parser for Alt-Ergo 🧚
  • --colors-in-output — prints output with colors 🌈
  • --frontend=FTD (absent=legacy) — picks the parsing and typing frontend 🧩
  • --no-forced-flush-in-output — prints output without the forced flush at the end of each print ✨
  • --no-formatting-in-output — don't use formatting rules in the output 🙅‍♀️
  • --no-headers-in-output — prints output without headers 🎀
  • --no-locs-in-answers — doesn't show the locations of goals when the solver prints its answers 📍
  • --parse-only — stops right after parsing 🛑
  • --prelude=VAL — adds a file that gets loaded as a prelude 📚 This one is cumulative, and the order of successive preludes is kept, so the order you list them in matters 🫶
  • --type-only — stops right after typing 🛑
  • --type-smt2 — stops right after SMT2 typing 🛑

✨ LIMIT OPTIONS ✨

  • --age-bound=AGE (absent=50) — sets the age limit bound ⏳
  • --fm-cross-limit=VAL (absent=10000) — skips the Fourier-Motzkin variable elimination steps that might produce more inequalities than the given limit 🚫 Unit eliminations always happen though, no matter what 💅
  • --timelimit-interpretation=SEC — sets the time limit to SEC seconds for model generation ⏱️ Also not supported on Windows, sorry bestie 😭
  • --timelimit-per-goal — the time limit you gave with --timelimit gets applied to each goal when a file has multiple goals 🎯 In that case, preprocessing time is counted separately from resolution time. This doesn't matter in GUI mode 🖥️

✨ INTERNAL OPTIONS ✨

  • --disable-weaks — stops the GC from collecting hashconsed data structures that aren't reachable 🗑️ This is useful if you want more determinism 🤓
  • --enable-assertions — turns on verification of some heavy invariants 🔍
  • --gc-policy=PLCY (absent=0) — sets the GC allocation policy 🧠 0 is next-fit, 1 is first-fit, and 2 is best-fit. Check out the Gc module of the OCaml distribution if you want more info 📖
  • --warning-as-error — treats warnings as errors ⚠️

✨ OUTPUT OPTIONS ✨

  • --interpretation=VAL (absent=0) — experimental counter-example generation 🕵️ The possible values are 1, 2, or 3, which compute an interpretation before returning Unknown, before instantiation (1), before every decision (2), or before every instantiation (3). A negative value (-1, -2, or -3) turns off interpretation display 🙈 Also, note that the --max-split limitation gets ignored during the model generation phase ✨

✨ CONTEXT OPTIONS ✨

  • --replay — replays a session that was saved in file_name.agr 🎬
  • --replay-all-used-context — replays using all the axioms and predicates saved in the .used files in the current directory 📂

✨ PROFILING OPTIONS ✨

  • --cumulative-time-profiling — records the time spent in called functions inside their callers ⏳
  • --profiling=DELAY — activates the profiling module at the given frequency 📊 Press Ctrl-C to switch between different views, and press Ctrl + AltGr + \ to exit 🦋
  • --profiling-plugin=PGN — uses the given profiling plugin 🔌

🐛 DEBUG OPTIONS 🐛

okay so these are the debug flags 🔍 each one prints debug info for one specific part of the solver. heads up though, bestie: they are not for checking internal consistency, so don't expect them to double-check the solver's homework 💅

  • --dac — turns on the debug flag for ac 🎀
  • --dadt — turns on the debug flag for ADTs
  • --darith — turns on the debug flag for Arith (without fm) ✨
  • --darrays — turns on the debug flag for arrays
  • --dbitv — turns on the debug flag for bitv
  • --dcc — turns on the debug flag for cc
  • --dcombine — turns on the debug flag for combine
  • --dconstr — turns on the debug flag for constructors
  • --debug-interpretation — turns on the debug flag for interpretation generation 🔮
  • --debug-unsat-core — replays the unsat-cores produced by --unsat-core. this one also implies --unsat-core, so you don't have to add it yourself 🫶
  • --dexplanations — turns on the debug flag for explanations
  • --dfm — turns on the debug flag for inequalities
  • --dfpa=INT (absent=0) — turns on the debug flag for floating-point 🌸
  • --dgc — prints some debug info about the GC's activity 🗑️
  • --dite — turns on the debug flag for ite
  • --dmatching=FLAG (absent=0) — turns on the debug flag for E-matching. 0 is disabled, 1 is light, and 2 is full 🦋
  • --dsat — turns on the debug flag for sat
  • --dsplit — turns on the debug flag for case-split analysis
  • --dsum — turns on the debug flag for Sum
  • --dtriggers — turns on the debug flag for triggers
  • --dtypes — turns on the debug flag for types
  • --dtyping — turns on the debug flag for typing (not the same as types, so don't mix them up 👀)
  • --duf — turns on the debug flag for uf
  • --duse — turns on the debug flag for use
  • --dwarnings — turns on the debug flag for warnings
  • --rule=TR (absent=none) — TR is one of parsing|typing|sat|cc|arith, and it prints the rule that got used on stderr 📣

🎀 COMMON OPTIONS 🎀

  • --help[=FMT] (default=auto) — shows this help in the format FMT. FMT has to be one of auto, pager, groff or plain. with auto, you get pager or plain whenever the TERM env var is dumb or undefined 💖
  • --version — shows the version info ✨

🚪 EXIT STATUS 🚪

alt-ergo exits with:

  • 0 — on success 🎉
  • 1 — on default errors
  • 123 — on indiscriminate errors reported on standard error
  • 124 — on command line parsing errors
  • 125 — on unexpected internal errors (bugs) 💀
  • 142 — on timeout errors ⏰

🐞 BUGS 🐞

You can open an issue on: https://github.com/OCamlPro/alt-ergo/issues 🫶

Or you can write to: alt-ergo@ocamlpro.com 💌

👑 AUTHORS 👑

✨ CURRENT AUTHORS ✨

  • Sylvain Conchon
  • Albin Coquereau
  • Guillaume Bury
  • Mattias Roux

🌈 ORIGINAL AUTHORS 🌈

  • Sylvain Conchon
  • Evelyne Contejean
  • Mohamed Iguernlala
  • Stephane Lescuyer
  • Alain Mebsout

That's the whole page, bestie 🦋 Alt-Ergo 2.4.3-free, manual page ALT-ERGO(1), all wrapped up and slay 💅🌸

📦 where this page comes from